MANIFOLDS OF DIFFERENTIABLE DENSITIES

被引:1
|
作者
Newton, Nigel J. [1 ]
机构
[1] Univ Essex, Sch Comp Sci & Elect Engn, Wivenhoe Pk, Colchester CO4 3SQ, Essex, England
关键词
Fisher-Rao Metric; Banach manifold; Frechet manifold; information geometry; non-parametric statistics; EXPONENTIAL STATISTICAL MANIFOLD; SPACE; GEOMETRY;
D O I
10.1051/ps/2018003
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
We develop a family of infinite-dimensional (non-parametric) manifolds of probability measures. The latter are defined on underlying Banach spaces, and have densities of class C-b(k) with respect to appropriate reference measures. The case k = infinity, in which the manifolds are modelled on Frechet spaces, is included. The manifolds admit the Fisher-Rao metric and, unusually for the non-parametric setting, Amari's alpha-covariant derivatives for all alpha is an element of R. By construction, they are C-infinity - embedded submanifolds of particular manifolds of finite measures. The statistical manifolds are dually (alpha = +/- 1) flat, and admit mixture and exponential representations as charts. Their curvatures with respect to the alpha-covariant derivatives are derived. The likelihood function associated with a finite sample is a continuous function on each of the manifolds, and the alpha-divergences are of class C-infinity.
引用
收藏
页码:19 / 34
页数:16
相关论文
共 50 条